reduction in PDF processing time
The problem
A growing backlog of PDFs was slowing down their entire product. Engineers manually keyed data from complex electrical drawings: a bottleneck killing delivery speed.
What I built
A Python + OCR pipeline that extracted structured data automatically from complex supplier PDFs: scanned images, inconsistent formats, mixed layouts across every document. Output fed directly into their Excel workflow via one-click macros. Delivered in 4 days.
Outcome
10,000+ database records populated automatically. Processing time dropped 99%. One person now manages what previously required a team.
"You are literally the tip of the spear here." Matt · Co-founder, SmartLineWork